Number of cluters of Kmedoids

Kmedoids_clusterN(dt)
## Loading required package: Matrix
## 
## Attaching package: 'Matrix'
## The following objects are masked from 'package:tidyr':
## 
##     expand, pack, unpack
Kmedoids_clusterN(dt, cluster = "Kmedoids-dr")
Kmedoids_gap(dt)
gap <- Kmedoids_gap(dt)
gap %>% group_by(representor, n_gap) %>% count()
gap %>% group_by(representor, n_gap, n_cluster) %>% count()
visualizeDistance(dt_orig, "ts-dr", "euclidean", "Kmedoids-dr")

inspect_silhouette(dt_orig, "ts-dr")
## Silhouette of 98 units in 2 clusters from pam(x = distance_mat, k = nl$other[[idx]]$n_cluster, diss = TRUE) :
##  Cluster sizes and average silhouette widths:
##        41        57 
## 0.2085589 0.3485292 
## Individual silhouette widths:
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
## -0.0196  0.1652  0.2863  0.2900  0.4040  0.5502 
## Silhouette of 98 units in 5 clusters from pam(x = distance_mat, k = nl$other[[idx]]$gap, diss = TRUE) :
##  Cluster sizes and average silhouette widths:
##        14        25        17        20        22 
## 0.2101185 0.3296237 0.2005931 0.0138171 0.2247882 
## Individual silhouette widths:
##     Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
## -0.27070  0.08895  0.21794  0.20218  0.32793  0.50213
visualizeDistance(dt_orig, "error-dr", "euclidean", "Kmedoids-dr")

inspect_silhouette(dt_orig, "error-dr")
## Silhouette of 98 units in 28 clusters from pam(x = distance_mat, k = nl$other[[idx]]$n_cluster, diss = TRUE) :
##  Cluster sizes and average silhouette widths:
##            3            6            6            2            6            2 
##  0.053431497  0.175802544 -0.018311458  0.209598984 -0.008925677  0.480756257 
##            3            3            4            4           13            3 
##  0.170764861  0.001260152  0.051239700  0.264549625  0.454430501  0.319260902 
##            1            3            1            3            7            3 
##  0.000000000  0.176022478  0.000000000  0.131458346 -0.026089228  0.282133661 
##            6            3            1            2            4            2 
##  0.640451563  0.265860729  0.000000000  0.875258185  0.200404800  0.797274173 
##            1            1            3            2 
##  0.000000000  0.000000000  0.363882808  0.898451930 
## Individual silhouette widths:
##     Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
## -0.15127  0.02426  0.16280  0.24835  0.44051  0.89959 
## Silhouette of 98 units in 2 clusters from pam(x = distance_mat, k = nl$other[[idx]]$gap, diss = TRUE) :
##  Cluster sizes and average silhouette widths:
##         53         45 
## 0.01100589 0.28603497 
## Individual silhouette widths:
##     Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
## -0.17119  0.02509  0.11531  0.13729  0.26949  0.44630
visualizeDistance(dt_orig, "accuracy", "euclidean")

Group Visualize

visualizeGroup(dt_orig, "error-dr", "euclidean", cluster= "Kmedoids-dr",names = dt_names)

visualizeGroup(dt_orig, "ts-dr", "euclidean", cluster= "Kmedoids-dr",names = dt_names)

visualizeGroup(dt_orig, "ts.features-dr", "euclidean", cluster= "Kmedoids-dr",names = dt_names)

visualizeGroup(dt_orig, "error.features-dr", "euclidean", cluster= "Kmedoids-dr",names = dt_names)

Overall statistics

avg_measure_fn(dt, metric = "rmsse") %>% arrange(bottom)

Overall rank mcb test

rank_compare(dt, filter_random = TRUE)
## Warning: Using an external vector in selections was deprecated in tidyselect 1.1.0.
## ℹ Please use `all_of()` or `any_of()` instead.
##   # Was:
##   data %>% select(measure)
## 
##   # Now:
##   data %>% select(all_of(measure))
## 
## See <https://tidyselect.r-lib.org/reference/faq-external-vector.html>.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
## Registered S3 method overwritten by 'quantmod':
##   method            from
##   as.zoo.data.frame zoo